All of lore.kernel.org
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	stable@vger.kernel.org, Russell King <rmk+kernel@armlinux.org.uk>,
	Robert Jarzmik <robert.jarzmik@free.fr>,
	"David S. Miller" <davem@davemloft.net>
Subject: [PATCH 4.4 41/73] net: smc91x: fix SMC accesses
Date: Wed, 28 Sep 2016 11:05:11 +0200	[thread overview]
Message-ID: <20160928090437.446120653@linuxfoundation.org> (raw)
In-Reply-To: <20160928090434.509091655@linuxfoundation.org>

4.4-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Russell King <rmk+kernel@armlinux.org.uk>


[ Upstream commit 2fb04fdf30192ff1e2b5834e9b7745889ea8bbcb ]

Commit b70661c70830 ("net: smc91x: use run-time configuration on all ARM
machines") broke some ARM platforms through several mistakes.  Firstly,
the access size must correspond to the following rule:

(a) at least one of 16-bit or 8-bit access size must be supported
(b) 32-bit accesses are optional, and may be enabled in addition to
    the above.

Secondly, it provides no emulation of 16-bit accesses, instead blindly
making 16-bit accesses even when the platform specifies that only 8-bit
is supported.

Reorganise smc91x.h so we can make use of the existing 16-bit access
emulation already provided - if 16-bit accesses are supported, use
16-bit accesses directly, otherwise if 8-bit accesses are supported,
use the provided 16-bit access emulation.  If neither, BUG().  This
exactly reflects the driver behaviour prior to the commit being fixed.

Since the conversion incorrectly cut down the available access sizes on
several platforms, we also need to go through every platform and fix up
the overly-restrictive access size: Arnd assumed that if a platform can
perform 32-bit, 16-bit and 8-bit accesses, then only a 32-bit access
size needed to be specified - not so, all available access sizes must
be specified.

This likely fixes some performance regressions in doing this: if a
platform does not support 8-bit accesses, 8-bit accesses have been
emulated by performing a 16-bit read-modify-write access.

Tested on the Intel Assabet/Neponset platform, which supports only 8-bit
accesses, which was broken by the original commit.

Fixes: b70661c70830 ("net: smc91x: use run-time configuration on all ARM machines")
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Tested-by: Robert Jarzmik <robert.jarzmik@free.fr>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 arch/arm/mach-pxa/idp.c                    |    3 -
 arch/arm/mach-pxa/xcep.c                   |    3 -
 arch/arm/mach-realview/core.c              |    3 -
 arch/arm/mach-sa1100/pleb.c                |    2 
 arch/blackfin/mach-bf561/boards/cm_bf561.c |    3 -
 arch/blackfin/mach-bf561/boards/ezkit.c    |    3 -
 drivers/net/ethernet/smsc/smc91x.c         |    7 +++
 drivers/net/ethernet/smsc/smc91x.h         |   65 ++++++++++++++++++++---------
 include/linux/smc91x.h                     |   10 ++++
 9 files changed, 73 insertions(+), 26 deletions(-)

--- a/arch/arm/mach-pxa/idp.c
+++ b/arch/arm/mach-pxa/idp.c
@@ -83,7 +83,8 @@ static struct resource smc91x_resources[
 };
 
 static struct smc91x_platdata smc91x_platdata = {
-	.flags = SMC91X_USE_32BIT | SMC91X_USE_DMA | SMC91X_NOWAIT,
+	.flags = SMC91X_USE_8BIT | SMC91X_USE_16BIT | SMC91X_USE_32BIT |
+		 SMC91X_USE_DMA | SMC91X_NOWAIT,
 };
 
 static struct platform_device smc91x_device = {
--- a/arch/arm/mach-pxa/xcep.c
+++ b/arch/arm/mach-pxa/xcep.c
@@ -120,7 +120,8 @@ static struct resource smc91x_resources[
 };
 
 static struct smc91x_platdata xcep_smc91x_info = {
-	.flags	= SMC91X_USE_32BIT | SMC91X_NOWAIT | SMC91X_USE_DMA,
+	.flags	= SMC91X_USE_8BIT | SMC91X_USE_16BIT | SMC91X_USE_32BIT |
+		  SMC91X_NOWAIT | SMC91X_USE_DMA,
 };
 
 static struct platform_device smc91x_device = {
--- a/arch/arm/mach-realview/core.c
+++ b/arch/arm/mach-realview/core.c
@@ -95,7 +95,8 @@ static struct smsc911x_platform_config s
 };
 
 static struct smc91x_platdata smc91x_platdata = {
-	.flags = SMC91X_USE_32BIT | SMC91X_NOWAIT,
+	.flags = SMC91X_USE_8BIT | SMC91X_USE_16BIT | SMC91X_USE_32BIT |
+		 SMC91X_NOWAIT,
 };
 
 static struct platform_device realview_eth_device = {
--- a/arch/arm/mach-sa1100/pleb.c
+++ b/arch/arm/mach-sa1100/pleb.c
@@ -45,7 +45,7 @@ static struct resource smc91x_resources[
 };
 
 static struct smc91x_platdata smc91x_platdata = {
-	.flags = SMC91X_USE_16BIT | SMC91X_NOWAIT,
+	.flags = SMC91X_USE_16BIT | SMC91X_USE_8BIT | SMC91X_NOWAIT,
 };
 
 static struct platform_device smc91x_device = {
--- a/arch/blackfin/mach-bf561/boards/cm_bf561.c
+++ b/arch/blackfin/mach-bf561/boards/cm_bf561.c
@@ -146,7 +146,8 @@ static struct platform_device hitachi_fb
 #include <linux/smc91x.h>
 
 static struct smc91x_platdata smc91x_info = {
-	.flags = SMC91X_USE_32BIT | SMC91X_NOWAIT,
+	.flags = SMC91X_USE_8BIT | SMC91X_USE_16BIT | SMC91X_USE_32BIT |
+		 SMC91X_NOWAIT,
 	.leda = RPC_LED_100_10,
 	.ledb = RPC_LED_TX_RX,
 };
--- a/arch/blackfin/mach-bf561/boards/ezkit.c
+++ b/arch/blackfin/mach-bf561/boards/ezkit.c
@@ -134,7 +134,8 @@ static struct platform_device net2272_bf
 #include <linux/smc91x.h>
 
 static struct smc91x_platdata smc91x_info = {
-	.flags = SMC91X_USE_32BIT | SMC91X_NOWAIT,
+	.flags = SMC91X_USE_8BIT | SMC91X_USE_16BIT | SMC91X_USE_32BIT |
+		 SMC91X_NOWAIT,
 	.leda = RPC_LED_100_10,
 	.ledb = RPC_LED_TX_RX,
 };
--- a/drivers/net/ethernet/smsc/smc91x.c
+++ b/drivers/net/ethernet/smsc/smc91x.c
@@ -2269,6 +2269,13 @@ static int smc_drv_probe(struct platform
 	if (pd) {
 		memcpy(&lp->cfg, pd, sizeof(lp->cfg));
 		lp->io_shift = SMC91X_IO_SHIFT(lp->cfg.flags);
+
+		if (!SMC_8BIT(lp) && !SMC_16BIT(lp)) {
+			dev_err(&pdev->dev,
+				"at least one of 8-bit or 16-bit access support is required.\n");
+			ret = -ENXIO;
+			goto out_free_netdev;
+		}
 	}
 
 #if IS_BUILTIN(CONFIG_OF)
--- a/drivers/net/ethernet/smsc/smc91x.h
+++ b/drivers/net/ethernet/smsc/smc91x.h
@@ -37,6 +37,27 @@
 #include <linux/smc91x.h>
 
 /*
+ * Any 16-bit access is performed with two 8-bit accesses if the hardware
+ * can't do it directly. Most registers are 16-bit so those are mandatory.
+ */
+#define SMC_outw_b(x, a, r)						\
+	do {								\
+		unsigned int __val16 = (x);				\
+		unsigned int __reg = (r);				\
+		SMC_outb(__val16, a, __reg);				\
+		SMC_outb(__val16 >> 8, a, __reg + (1 << SMC_IO_SHIFT));	\
+	} while (0)
+
+#define SMC_inw_b(a, r)							\
+	({								\
+		unsigned int __val16;					\
+		unsigned int __reg = r;					\
+		__val16  = SMC_inb(a, __reg);				\
+		__val16 |= SMC_inb(a, __reg + (1 << SMC_IO_SHIFT)) << 8; \
+		__val16;						\
+	})
+
+/*
  * Define your architecture specific bus configuration parameters here.
  */
 
@@ -55,10 +76,30 @@
 #define SMC_IO_SHIFT		(lp->io_shift)
 
 #define SMC_inb(a, r)		readb((a) + (r))
-#define SMC_inw(a, r)		readw((a) + (r))
+#define SMC_inw(a, r)							\
+	({								\
+		unsigned int __smc_r = r;				\
+		SMC_16BIT(lp) ? readw((a) + __smc_r) :			\
+		SMC_8BIT(lp) ? SMC_inw_b(a, __smc_r) :			\
+		({ BUG(); 0; });					\
+	})
+
 #define SMC_inl(a, r)		readl((a) + (r))
 #define SMC_outb(v, a, r)	writeb(v, (a) + (r))
+#define SMC_outw(v, a, r)						\
+	do {								\
+		unsigned int __v = v, __smc_r = r;			\
+		if (SMC_16BIT(lp))					\
+			__SMC_outw(__v, a, __smc_r);			\
+		else if (SMC_8BIT(lp))					\
+			SMC_outw_b(__v, a, __smc_r);			\
+		else							\
+			BUG();						\
+	} while (0)
+
 #define SMC_outl(v, a, r)	writel(v, (a) + (r))
+#define SMC_insb(a, r, p, l)	readsb((a) + (r), p, l)
+#define SMC_outsb(a, r, p, l)	writesb((a) + (r), p, l)
 #define SMC_insw(a, r, p, l)	readsw((a) + (r), p, l)
 #define SMC_outsw(a, r, p, l)	writesw((a) + (r), p, l)
 #define SMC_insl(a, r, p, l)	readsl((a) + (r), p, l)
@@ -66,7 +107,7 @@
 #define SMC_IRQ_FLAGS		(-1)	/* from resource */
 
 /* We actually can't write halfwords properly if not word aligned */
-static inline void SMC_outw(u16 val, void __iomem *ioaddr, int reg)
+static inline void __SMC_outw(u16 val, void __iomem *ioaddr, int reg)
 {
 	if ((machine_is_mainstone() || machine_is_stargate2() ||
 	     machine_is_pxa_idp()) && reg & 2) {
@@ -405,24 +446,8 @@ smc_pxa_dma_insw(void __iomem *ioaddr, s
 
 #if ! SMC_CAN_USE_16BIT
 
-/*
- * Any 16-bit access is performed with two 8-bit accesses if the hardware
- * can't do it directly. Most registers are 16-bit so those are mandatory.
- */
-#define SMC_outw(x, ioaddr, reg)					\
-	do {								\
-		unsigned int __val16 = (x);				\
-		SMC_outb( __val16, ioaddr, reg );			\
-		SMC_outb( __val16 >> 8, ioaddr, reg + (1 << SMC_IO_SHIFT));\
-	} while (0)
-#define SMC_inw(ioaddr, reg)						\
-	({								\
-		unsigned int __val16;					\
-		__val16 =  SMC_inb( ioaddr, reg );			\
-		__val16 |= SMC_inb( ioaddr, reg + (1 << SMC_IO_SHIFT)) << 8; \
-		__val16;						\
-	})
-
+#define SMC_outw(x, ioaddr, reg)	SMC_outw_b(x, ioaddr, reg)
+#define SMC_inw(ioaddr, reg)		SMC_inw_b(ioaddr, reg)
 #define SMC_insw(a, r, p, l)		BUG()
 #define SMC_outsw(a, r, p, l)		BUG()
 
--- a/include/linux/smc91x.h
+++ b/include/linux/smc91x.h
@@ -1,6 +1,16 @@
 #ifndef __SMC91X_H__
 #define __SMC91X_H__
 
+/*
+ * These bits define which access sizes a platform can support, rather
+ * than the maximal access size.  So, if your platform can do 16-bit
+ * and 32-bit accesses to the SMC91x device, but not 8-bit, set both
+ * SMC91X_USE_16BIT and SMC91X_USE_32BIT.
+ *
+ * The SMC91x driver requires at least one of SMC91X_USE_8BIT or
+ * SMC91X_USE_16BIT to be supported - just setting SMC91X_USE_32BIT is
+ * an invalid configuration.
+ */
 #define SMC91X_USE_8BIT (1 << 0)
 #define SMC91X_USE_16BIT (1 << 1)
 #define SMC91X_USE_32BIT (1 << 2)

  parent reply	other threads:[~2016-09-28  9:32 UTC|newest]

Thread overview: 69+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <CGME20160928090623uscas1p1076bd85a3fd981ed5a1284f5bebb1bbf@uscas1p1.samsung.com>
2016-09-28  9:04 ` [PATCH 4.4 00/73] 4.4.23-stable review Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 01/73] include/linux/kernel.h: change abs() macro so it uses consistent return type Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 02/73] Fix build warning in kernel/cpuset.c Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 03/73] reiserfs: fix "new_insert_key may be used uninitialized ..." Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 09/73] crypto: arm64/aes-ctr - fix NULL dereference in tail processing Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 10/73] crypto: arm/aes-ctr " Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 11/73] crypto: skcipher - Fix blkcipher walk OOM crash Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 12/73] crypto: echainiv - Replace chaining with multiplication Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 13/73] ocfs2/dlm: fix race between convert and migration Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 14/73] ocfs2: fix start offset to ocfs2_zero_range_for_truncate() Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 15/73] kbuild: Do not run modules_install and install in paralel Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 16/73] Makefile: revert "Makefile: Document ability to make file.lst and file.S" partially Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 17/73] tools: Support relative directory path for O= Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 18/73] kbuild: forbid kernel directory to contain spaces and colons Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 19/73] Kbuild: disable maybe-uninitialized warning for CONFIG_PROFILE_ALL_BRANCHES Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 20/73] gcov: disable -Wmaybe-uninitialized warning Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 21/73] Disable "maybe-uninitialized" warning globally Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 23/73] Makefile: Mute warning for __builtin_return_address(>0) for tracing only Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 24/73] net: caif: fix misleading indentation Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 26/73] [media] am437x-vfpe: fix typo in vpfe_get_app_input_index Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 27/73] ath9k: fix misleading indentation Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 28/73] iwlegacy: avoid warning about missing braces Greg Kroah-Hartman
2016-09-28  9:04   ` [PATCH 4.4 29/73] Staging: iio: adc: fix indent on break statement Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 30/73] nouveau: fix nv40_perfctr_next() cleanup regression Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 31/73] megaraid: fix null pointer check in megasas_detach_one() Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 32/73] bonding: Fix bonding crash Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 33/73] Revert "af_unix: Fix splice-bind deadlock" Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 34/73] af_unix: split u->readlock into two: iolock and bindlock Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 35/73] vti: flush x-netns xfrm cache when vti interface is removed Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 36/73] net/irda: handle iriap_register_lsap() allocation failure Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 37/73] tipc: fix NULL pointer dereference in shutdown() Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 38/73] net/mlx5: Added missing check of msg length in verifying its signature Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 39/73] net: dsa: bcm_sf2: Fix race condition while unmasking interrupts Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 40/73] Revert "phy: IRQ cannot be shared" Greg Kroah-Hartman
2016-09-28  9:05   ` Greg Kroah-Hartman [this message]
2016-09-28  9:05   ` [PATCH 4.4 42/73] bridge: re-introduce fix parsing of MLDv2 reports Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 43/73] pwm: Mark all devices as "might sleep" Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 44/73] autofs races Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 45/73] autofs: use dentry flags to block walks during expire Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 46/73] xfs: prevent dropping ioend completions during buftarg wait Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 47/73] fsnotify: add a way to stop queueing events on group shutdown Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 48/73] fanotify: fix list corruption in fanotify_get_response() Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 49/73] fix fault_in_multipages_...() on architectures with no-op access_ok() Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 50/73] mtd: maps: sa1100-flash: potential NULL dereference Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 51/73] mtd: pmcmsp-flash: Allocating too much in init_msp_flash() Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 52/73] power: reset: hisi-reboot: Unmap region obtained by of_iomap Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 53/73] fix memory leaks in tracing_buffers_splice_read() Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 54/73] tracing: Move mutex to protect against resetting of seq data Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 55/73] mm: delete unnecessary and unsafe init_tlb_ubc() Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 56/73] can: flexcan: fix resume function Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 57/73] nl80211: validate number of probe response CSA counters Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 58/73] btrfs: ensure that file descriptor used with subvol ioctls is a dir Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 59/73] i2c-eg20t: fix race between i2c init and interrupt enable Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 60/73] i2c: qup: skip qup_i2c_suspend if the device is already runtime suspended Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 61/73] MIPS: Fix pre-r6 emulation FPU initialisation Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 63/73] MIPS: vDSO: Fix Malta EVA mapping to vDSO page structs Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 64/73] MIPS: Remove compact branch policy Kconfig entries Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 65/73] MIPS: Avoid a BUG warning during prctl(PR_SET_FP_MODE, ...) Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 66/73] MIPS: Add a missing ".set pop" in an early commit Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 67/73] MIPS: paravirt: Fix undefined reference to smp_bootstrap Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 68/73] PM / hibernate: Restore processor state before using per-CPU variables Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 69/73] PM / hibernate: Fix rtree_next_node() to avoid walking off list ends Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 70/73] power_supply: tps65217-charger: fix missing platform_set_drvdata() Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 71/73] power: supply: max17042_battery: fix model download bug Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 72/73] qxl: check for kmap failures Greg Kroah-Hartman
2016-09-28  9:05   ` [PATCH 4.4 73/73] hostfs: Freeing an ERR_PTR in hostfs_fill_sb_common() Greg Kroah-Hartman
2016-09-28 16:45   ` [PATCH 4.4 00/73] 4.4.23-stable review Shuah Khan
2016-09-28 22:43   ` Guenter Roeck
     [not found]   ` <57ec0f9e.07ddc20a.146f7.4be3@mx.google.com>
2016-09-29  9:01     ` Greg Kroah-Hartman

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20160928090437.446120653@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=davem@davemloft.net \
    --cc=linux-kernel@vger.kernel.org \
    --cc=rmk+kernel@armlinux.org.uk \
    --cc=robert.jarzmik@free.fr \
    --cc=stable@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.